Close Window

Challenges of a coupled climate-biosphere model to reproduce vegetation dynamics in Amazonia

Mônica Carneiro Alves Senna, Universidade Federal de Viçosa, (Presenting)
Marcos Heil Costa, Universidade Federal de Viçosa,

We investigate how well a fully coupled atmosphere-biosphere model, CCM3-IBIS, can reproduce vegetation dynamics in Amazonia. We conduct an experiment with three ensembles of the global climate, with fixed climatological sea-surface temperatures and with dynamical vegetation, for a period of 30 years. Two climate variables, precipitation and incident solar radiation, and four vegetation dynamics variables, aboveground live biomass (AGLB), leaf area index (LAI), net primary production (NPP) and land cover, are compared to observations. Precipitation is compared with data from NCEP/NCAR reanalysis, ERA-40 project, CMAP (CPC Merged Analysis of Precipitation), GPCP (Global Precipitation Climatology Project), TRMM (Tropical Rainfall Measuring Mission), CRU, Legates and Willmott (1990) and Leemans and Cramer (1990) databases. Incident solar radiation is compared with GOES algorithm GL1.2. Both climate variables are analyzed for the entire South America. AGLB is compared with the recent map by Saatchi et al. (2007). LAI and NPP are compared with in situ measurements done by LBA researchers. Land cover is compared with land use map from SAGE (Center for Sustainability and the Global Environment). The vegetation variables are analyzed for Amazonia Tropical Forest. Results indicate that the Amazon climate (annual mean and seasonality) is extremely well simulated for both precipitation and incident solar radiation. Vegetation cover patterns reproduce well the observed patterns. Although the climate variables are well represented, there are some differences in the AGLB simulated. These errors in simulated AGLB are probably because the allocation coefficient and resident time of wood, leave and root need to be better represented.

Science Theme:  LC (Land Use and Land Cover Change)

Presentation Type:  Poster (view presentation (217 KB))

Abstract ID: 44

Close Window